Incorporation of Substitutionally Labile - American Chemical Society

Mar 6, 2008 - emu · Oe/mol). Reactions of [VIII(CN)6]3- with [MII(NCMe)6]2+ (M ) Fe, Co, Ni) in MeCN lead to either partial (M. ) Co) or complete (M ...
0 downloads 0 Views 581KB Size
Download PDF
Inorg. Chem. 2008, 47, 2526-2533

Incorporation of Substitutionally Labile [VIII(CN)6]3- into Prussian Blue Type Magnetic Materials Kendric J. Nelson and Joel S. Miller* Department of Chemistry, UniVersity of Utah, 315 South 1400 East, Room 2020, Salt Lake City, Utah 84112-0850 Received September 18, 2007

A Prussian blue (PB) type material containing hexacyanovanadate(III), MnII1.5[VIII(CN)6] · 0.30MeCN (1), was formed from the reaction of [VIII(CN)6]3- with [Mn(NCMe)6]2+ in MeCN. This new material exhibits ferrimagnetic spin- or cluster-glass behavior below a Tc of 12 K with observed magnetic hysteresis at 2 K (Hcr ) 65 Oe and Mrem ) 730 emu · Oe/mol). Reactions of [VIII(CN)6]3- with [MII(NCMe)6]2+ (M ) Fe, Co, Ni) in MeCN lead to either partial (M ) Co) or complete (M ) Fe, Ni) linkage isomerization, resulting in compounds of FeII0.5VIII[FeII(CN)6] · 0.85MeCN (2), (NEt4)0.10CoII1.5-aVIIa[CoIII(CN)6]a[VIII(CN)6]1-a(BF4)0.10 · 0.35MeCN (3), and (NEt4)0.20VIII[NiII(CN)4]1.6 · 0.10MeCN (4) compositions. Compounds 2–4 do not magnetically order as a consequence of diamagnetic cyanometalate anions being present, i.e., [FeII(CN)6]4-, [CoIII(CN)6]3-, and [NiII(CN)4]2-. Incorporation of [VIII(CN)6]3- into PB-type materials is synthetically challenging because of the lability of the cyanovanadate(III) anion.

Introduction The synthesis of Prussian blue (PB) and its analogues, leading to new molecule-based magnetic materials, has been a contemporary research focus.1 Molecule-based magnets have potential uses in many areas including magnetic shielding2 and spintronic memory storage devices.3 Prussian blue analogues (PBAs) are generally synthesized by mixing a substitutionally labile [M(Solvent)x]n+ cation with a * To whom correspondence should be addressed. E-mail: [email protected]. (1) (a) Mallah, T.; Thiébaut, S.; Verdaguer, M.; Veillet, P. Science 1993, 262, 1554. (b) Ferlay, S.; Mallah, T.; Ouahes, R.; Veillet, P.; Verdaguer, M. Nature 1995, 378, 701. (c) Buschmann, W. E.; Paulson, S. C.; Wynn, C. M.; Girtu, M.; Epstein, A. J.; White, H. S.; Miller, J. S. AdV. Mater. 1997, 9, 645. (d) Buschmann, W. E.; Paulson, S. C.; Wynn, C. M.; Girtu, M.; Epstein, A. J.; White, H. S.; Miller, J. S. Chem. Mater. 1998, 10, 1386. (e) Dujardin, E.; Ferlay, S.; Phan, X.; Desplanches, C.; Moulin, C. C. D.; Sainctavit, P.; Baudelet, F.; Dartyge, E.; Veillet, P.; Verdaguer, M. J. Am. Chem. Soc. 1998, 120, 11347. (f) Ferlay, S.; Mallah, T.; Ouahes, R.; Veillet, P.; Verdaguer, M. Inorg. Chem. 1999, 38, 229. (g) Verdaguer, M.; Bleuzen, A.; Marvaud, V.; Vaissermann, J.; Seuleiman, M.; Desplanches, C.; Scuiller, A.; Train, C.; Garde, R.; Gelly, G.; Lomenech, C.; Rosenman, I.; Veillet, P.; Cartier, C.; Villain, F. Coord. Chem. ReV. 1999, 190–192, 1023. (h) Verdaguer, M.; Bleuzen, A.; Train, C.; Garde, R.; Fabrizi de Biani, F.; Desplanches, C. Philos. Trans. R. Soc. London, Ser. A 1999, 357, 2959. (i) Hashimoto, K.; Ohkoshi, S. Philos. Trans. R. Soc. London, Ser. A 1999, 357, 2977. (j) HatlevikØ.; Buschmann, W. E.; Zhang, J.; Manson, J. L.; Miller, J. S. AdV. Mater. 1999, 11, 914. (k) Holmes, S. M.; Girolami, G. S. J. Am. Chem. Soc. 1999, 121, 5593. (l) Verdaguer, M.; Girolami, G. S. In Magnetisms Molecules to Materials; Miller, J. S., Drillon, M., Eds.; Wiley-VCH: Weinheim, Germany, 2005; Vol. 5, pp 283–346.

2526 Inorganic Chemistry, Vol. 47, No. 7, 2008

substitutionally inert [M′(CN)6]m- anion in aqueous media to form AaMb[M′(CN)6]c · zSolvent, which adopts a facecentered-cubic (fcc) structure with unit cell length a ) 10.4 ( 0.5 Å.4,5 Because of the electronic and spin-state differences available to both MII and M′, PBAs can be tailored to possess numerous metal ions that vary in the charge and number of spins per metal site and thus can lead to a variety of magnetic behaviors.1 Incorporating the cyanovanadate(III) anion into PB materials has proven challenging because it is hydrolytically unstable in aqueous media.6 However, with the advent of the nonaqueous soluble cyanovanadate(III) anion in (NEt4)3[VIII(CN)6]7, it was anticipated that incorporation of [VIII(CN)6]3- into PB materials by nonaqueous routes was possible. Initial attempts to synthesize PB materials possessing hexacyanovanadate(III) were unsuccessful, because while (2) (a) Landee, C. P.; Melville, D.; Miller, J. S. In NATO ARW Molecular Magnetic Materials; Kahn, O., Gatteschi, D., Miller, J. S., Palacio, F., Eds.; Kluwer Academic Publishers: London, 1991; Vol. E198, p 395. (b) Morin, B. G.; Hahm, C.; Epstein, A. J.; Miller, J. S. J. Appl. Phys. 1994, 75, 5782. (c) Miller, J. S. AdV. Mater. 1994, 6, 322. (3) Prigodin, V. N.; Raju, N. P.; Pokhodnya, K. I.; Miller, J. S.; Epstein, A. J. AdV. Mater. 2002, 14, 1230. (4) Verdaguer, M.; Girolami, G. S. In MagnetismsMolecules to Materials; Miller, J. S., Drillon, M., Eds.; Wiley-VCH: Weinheim, Germany, 2005; Vol. 5, pp 283–346. (5) (a) Ludi, A.; Güdel, H. U. Struct. Bonding (Berlin) 1973, 14, 1. (b) Dunbar, K. R.; Heintz, R. A. Prog. Inorg. Chem. 1997, 45, 283. (6) Levenson, R. A.; Dominguez, R. J. G.; Willis, M. A.; Young, F. R., III. Inorg. Chem. 1974, 13, 2761.

10.1021/ic701845p CCC: $40.75

 2008 American Chemical Society Published on Web 03/06/2008

Incorporation of Substitutionally Labile [VIII(CN)6]3-

the reaction of (NEt4)3[VIII(CN)6] with [CrII(NCMe)4](BF4)2 in MeCN formed Cr1.5[V(CN)6] · zMeCN, electron transfer between CrII and VIII occurs and the product is best formulated as CrII0.5CrIII[VII(CN)6] · zMeCN.7 An exception was the formation of CrIII[VIII(CN)6] using the unconventional [CrIII(NCMe)6]3+ nonaqueous labile source of CrIII and [VIII(CN)6]3-.8 With the goal to prepare [VIII(CN)6]3--based PB-type magnets, the reaction of [VIII(CN)6]3- with [MII(NCMe)6]2+ was investigated in nonaqueous media, and herein we report the reactions involving MII (M ) Mn, Fe, Co, Ni) and the formation of a new [VIII(CN)6]3--based magnet. Experimental Section All manipulations were carried out under a dry N2 atmosphere ( Mn >> Cr > Fe ∼ Co), it is more likely the latter linkage has been formed because of the stability of the [CoIII(CN)6]3- anion.28 Hence, 3 has both CoII-NtC-VIII and CoIII-CtN-VII linkages present in the structure, which can be assigned to the 2121 and 2161 cm-1 absorptions, respectively. It is determined that the ratio of these two linkages cannot be directly obtained from the IR spectra; therefore, further insight is needed from the magnetic analyses to draw any further conclusions on the chemical formulation of this compound (vide infra). For now, we can represent this compound with the generic formula (NEt4)0.10CoII1.5-aVIIa[CoIII(CN)6]a[VIII(CN)6]1-a(BF4)0.10 · 0.35MeCN (a < 1). Compound 4 exhibits a strong νCN absorption at 2168 cm-1 (hwhh: 28 cm-1). Again, this absorption is shifted to higher energy with respect to [VIII(CN)6]3-, and it is an average of ∼43 cm-1 higher than the observed νCN absorption in compounds 1 and 3, which were assigned to MII-NtC-VIII linkages. Because of this large shift for the NiII compound, isomerization of the desired NiII-NtC-VIII linkages are considered. Linkage isomerization in 4 would result in the formation of NiII-CtN-VIII. Cyanonickelate(II) complexes can exist as tetracyanonickelates, and in some cases pentacyanonickelates. In the solid state, [NiII(CN)4]2- has been reported to have a νCN absorption occurring at 2142 cm-1,29 whereas [NiII(CN)5]3- has four observed νCN bands at 2076, (23) (a) Alexander, J. J.; Gray, H. B. J. Am. Chem. Soc. 1967, 89, 3356. (b) Griffith, W. P.; Lane, J. R. J. Chem. Soc., Dalton Trans. 1972, 158. (c) White, D. A.; Solodar, A. J.; Baizer, M. M. Inorg. Chem. 1972, 11, 2160. (d) Brown, L. D.; Raymond, K. N. Inorg. Chem. 1975, 14, 2590. (24) White, D. A.; Solodar, A. J.; Baizer, M. M. Inorg. Chem. 1972, 11, 2160. (25) Beauvais, L. G.; Long, J. R. J. Am. Chem. Soc. 2002, 124, 12096. (26) Jones, L. H. Inorg. Chem. 1963, 2, 777. (27) (a) Ludi, A.; Güdel, H. U. HelV. Chim. Acta 1968, 51, 2006. (b) Shiver, D. F.; Brown, D. B. Inorg. Chem. 1969, 8, 42. (c) Griebler, W.-D.; Babel, D. Z. Naturforsch. 1982, 37b, 832. (d) Buchold, D. H. M.; Feldmann, C. Chem. Mater. 2007, 19, 2276. (28) (a) Taube, H. Chem. ReV. 1952, 50, 69. (b) Basolo, F.; Pearson, R. G. Mechanism of Inorganic Reactions: A Study of Metal Complexes in Solution, 2nd ed.; John Wiley and Sons Inc.: New York, 1967; p 157. (c) Atwood, J. D. Inorganic and Organometallic Reaction Mechanisms, 2nd ed.; Wiley-VCH:New York, 1997; p 84. (29) Kubas, G. J.; Jones, L. H. Inorg. Chem. 1974, 13, 2816.

2100, 2114, and 2127 cm-1.30 It was also reported that NiII(CN)2 · 1.5H2O,31 an extended network solid, was shown to have the presence of NiII-CtN-NiII linkages,8,32 and the νCN absorption for this binding motif was reported to be 2170 cm-1.33 In addition, numerous other extended network materials containing the four-coordinate [NiII(CN)4]2- with its cyanides bridging to another metal have an observable range of νCN absorptions from 2143 to 2195 cm-1.34 Hence, the 2168 cm-1 νCN absorption is more consistent with the cyanides being C-bound to NiII than the VIII forming NiII-CtN-VIII linkages. Therefore, 3 undergoes linkage isomerization, forming (NEt4)0.20VIII[NiII(CN)4]1.6 · 0.10MeCN. This formulation is further confirmed by magnetic analysis (vide infra). Compounds 1–4 also have small amounts of both coordinated (νCN ) 2286 ( 3 and 2313 ( 2 cm-1) and noncoordinated MeCN (νCN ) 2252 cm-1). Upon exposure to air, these absorptions rapidly disappear and a simultaneous growth of νOH absorptions are observed. Thus, MeCN can be easily displaced by water upon exposure to moisture, and it is likely that MeCN in these materials is on the surface and/or in the interstitial lattice sites, as previously reported.35 The IR spectra for these compounds also show trace amounts of the NEt4+ cation as well as anion (SbF6- in 1 or BF4- in 2-4) that were minimized, but not eliminated, by thoroughly washing the solids with MeCN. The samples chosen for analyses were verified to have minimal amounts of these ions by IR analyses. Thermal Stability. The thermal properties of compounds 1–4 were analyzed by TGA/MS (Figures S2-S5 in the Supporting Information). In general, each compound showed loss of MeCN and cyanide at 50 °C, followed by a continuous gradual mass loss up to 600 °C, all resulting in black residues. Because of this gradual mass loss, attributed to simultaneous solvent loss and decomposition, the temperatures at which these events occur are indistinguishable. Nevertheless, the total mass losses were determined to be 26.0, 56.7, 49.7, and 46.5%. Overall, these materials are not very thermally stable because of their facile loss of MeCN and cyanide at relatively low temperatures (∼50 °C); therefore, all materials were stored at -25 °C, and exposure (30) Raymond, K. N.; Corfield, P. W. R.; Ibers, J. A. Inorg. Chem. 1968, 7, 1362. (31) Ludi, A.; Hügi, R. HelV. Chim. Acta 1967, 50, 1283. (32) The NiII-CtN-NiII linkages come from square-planar [NiII(CN)4]2sites bridging through the cyanide to octahedral [NiII(NC)4(OH2)2]2sites. (33) (a) El-Sayed, M. F. A.; Sheline, R. K. J. Inorg. Nucl. Chem. 1958, 6, 187. (b) Ludi, A.; Hügi, R. HelV. Chim. Acta 1968, 51, 349. (34) (a) Özbay, A.; Yurdakul, S¸. J. Mol. Struct. 1995, 349, 13. (b) Karacan, N.; Kantarcý, Z.; Akyüz, S. Spectrochim. Acta, Part A 1996, 52, 771. (c) Yurdakul, S¸; Güllüogˇlu, M. T.; Küçükgüldal, D.; Taþdelen, M. J. Mol. Struct. 1997, 408–409, 319. (d) Yurdakul, S¸; Bahat, M. J. Mol. Struct. 1997, 412, 97. (e) Kasap, E.; Özbay, A.; Özçelik, S. Spectrosc. Lett. 1997, 30, 491. (f) Akyüz, S. J. Mol. Struct. 1998, 449, 23. (g) Yurdakul, S¸.; Yilmaz, C. Vib. Spectrosc. 1999, 21, 127. (h) Serbakan, T. R.; Sagˇlam, S.; Kasap, E.; Kantarci, Z. Vib. Spectrosc. 2000, 24, 249. (i) Bahat, M.; Yurdakul, Sˇ. Spectrochim. Acta, Part A 2002, 58, 933. (j) Akyuz, S. J. Mol. Struct. 2003, 651–653, 541. (k) Chen, X.Y.; Shi, W.; Xia, J.; Cheng, P.; Zhao, B.; Song, H.-B.; Wang, H.-G.; Yan, S.-P.; Liao, D.-Z.; Jiang, Z.-H. Inorg. Chem. 2005, 44, 4263. (l) Yeung, W.-F.; Kwong, H.-K.; Lau, T.-C.; Gao, S.; Szeto, L.; Wong, W.-T. Polyhedron 2006, 25, 1256. (35) Buschmann, W. E.; Miller, J. S. Inorg. Chem. 2000, 39, 2411.

Inorganic Chemistry, Vol. 47, No. 7, 2008

2529

Nelson and Miller Table 2. Summary of Magnetic Properties for 1–4 1

2

3

4

5.29 2.20 2.35 0.95 χTobs, emu · K/mola χTcalc, emu · K/molb 7.39 2.66 3.14c 0.83 5.54 2.61 2.99c χT(T-θ)calc, emu · K/mol 0.82 θ, K (Curie–Weiss) -4 -6 -15 -3 θ, K (Néel) -100 TN, K (dc) 90 Tonset, K (ZFC) 35 Tb, K (ZFC/FC) 10 Tc, K (ac) 12 Tf, K (ac) 11.5 φ (ac) 0.029 Hcr, Oe 65 M, emu · Oe/mol, (at 90 kOe) 21 000 8000 10 800 7180 Mrem, emu · Oe/mol 780 a At room temperature. b Calculated spin-only values. c These χT values are for a ) 1 in (NEt4)0.10CoII1.5-aVIIa[CoIII(CN)6]a[VIII(CN)6]1-a(BF4)0.10 · 0.35MeCN.

to ambient temperatures was minimized. Drying of the samples before analyses was performed at room temperature to avoid any thermal decomposition at elevated temperatures. PXRD. Generally, PBAs adopt a fcc structure with a unit cell lattice constant (a) ranging from 9.9 to 10.9 Å;4,5 however, the diffraction patterns of 1–4 showed weak broad peaks indicative of being completely amorphous, and none of the patterns could be indexed to a fcc structure. Because of their amorphous nature, these materials are structurally disordered, and this results in small sample-to-sample variations in both the structural and magnetic properties. The synthesis of more crystalline samples was attempted by varying the addition rate in the synthesis; however, all samples were concluded to be amorphous regardless of the synthetic procedure. Magnetic Susceptibility. The temperature dependencies of the susceptibility, χ(T), of 1–4 were measured at 500 Oe between 5 and 300 K. From these data, it was concluded that only compound 1 exhibits bulk magnetic ordering. The magnetic properties of 1–4 are summarized in Table 2 and are individually discussed below. Compound 1. 1 has a room temperature χT value of 5.29 emu · K/mol that is substantially lower than the spin-only value of 7.39 emu · K/mol for S ) 1 and g ) 1.82 (VIII)7 and S ) 5/2 and g ) 2.00 (MnII).36 Above 250 K, χ(T) was fit to the Curie–Weiss expression χ ∝ (T - θ)-1, with θ ) -4 K, suggesting that antiferromagnetic coupling dominates the short-range exchange.37 As the temperature is lowered, χT(T) gradually increases until ∼100 K, increasing more rapidly to a maximum of 163 emu · K/mol at 28 K and then finally decreasing rapidly down to 43.3 emu · K/mol at 5 K (Figure 1). The rise in χT(T) to a maximum is consistent with the onset of magnetic ordering at ∼70 K from an extrapolation of the steepest slope in χT(T) to zero. The exact ordering temperature, Tc, is determined via alternate methods. Ferrimagnetic ordering is anticipated because of the negative θ value determined from the Curie–Weiss fit. Also consistent with ferrimagnetic ordering is the expected minimum in (36) Watanabe, T. J. Phys. Soc. Jpn. 1961, 16, 1131. (37) Caution should be taken with the Curie-Weiss fit of compound 1 because of the small temperature range that was used to fit the χ-1(T) data (250-300 K).

2530 Inorganic Chemistry, Vol. 47, No. 7, 2008

Figure 1. χT(T) (b) and χ-1(T) (O) for 1. The fit to Néel’s hyperbolic equation (eq 1) is displayed with a solid line. The sample was cooled in a zero applied field to 5 K, and data were measured upon warming in a 500 Oe applied field.

χT(T) above Tc. This results from short-range antiferromagnetic correlations and causes a cancelation of spins as the temperature is lowered.38,39 This expected minimum, however, is not observed for 1. The absence of this minimum may be explained by a lack of thermal energy at room temperature needed to completely compensate for magnetic coupling energy, which causes thermal randomization of the electron spins, and the χT(T) data will likely have a minimum occurring at elevated temperatures. Hence, the reported θ value is an effective θ and not the true value. For materials exhibiting an onset of ferrimagnetic ordering, the θ value can also be determined from Néel’s theory39 (eq 1), where C and θ are the Curie and Weiss constants, respectively, and θ′ and ζ are proportional to ηAηBC(ηA ηB) and ηAηBC (where ηi is the fractional occupancy of each sublattice site), respectively. Above 110 K, χ-1(T) for 1 was fit to eq 1 with a Néel temperature, TN, of 90 K and a θ value of -100 K. Using this θ value, χT(T - θ)calc was calculated to be 5.54 emu · K/mol, which is more consistent with the observed room temperature χT value of 5.29 emu · K/ mol. The less than ideal hyperbolic shape of the fit to the χ-1(T) data suggests the presence of several spin domains. Inside these domains, the spins are well ordered just above 90 K; however, these domains do not correlate throughout the bulk material until a lower temperature is obtained. T-θ ζ (1) C T - θ′ Low-field M(T) measurements are used to determine key information about an ordering system, and low-field MZFC(T)/ MFC(T) (ZFC and FC represent zero-field-cooled and fieldcooled, respectively) of 1 shows the onset of magnetic ordering, Tonset, below 35 K (by extrapolation of the steepest slope to T where M ) 0) and a blocking temperature, Tb, of 10 K (Figure 2). Tb is the temperature where the MZFC(T) and MFC(T) magnetizations converge, and below this, the χ-1 )

(38) (a) Stumpf, H. O.; Pei, Y.; Kahn, O.; Sletten, J.; Renard, J. P. J. Am. Chem. Soc. 1993, 115, 6738. (b) Stumpf, H. O.; Ouahab, L.; Pei, Y.; Bergerat, P.; Kahn, O. J. Am. Chem. Soc. 1994, 116, 3866. (39) (a) Néel, L. Ann. Phys. 1948, 3, 137. (b) Smart, J. S. Am. J. Phys. 1955, 23, 356.

Incorporation of Substitutionally Labile [VIII(CN)6]3-

Figure 4. Magnetic hysteresis curve for 1. Figure 2. 5 Oe MZFC (b) and MFC (O) data for 1.

Figure 3. Observed χ′(T) (closed symbols) and χ″(T) (open symbols) ac responses for 1 at 33 (O), 100 (0), 333 (]), and 1000 (4) Hz in zero applied field.

temperature is irreversibile (or hysteresis) in these magnetizations as a result of entering the magnetically ordered state. The decrease in magnetization below the peak in the MZFC(T)/ MFC(T) data is consistent with this sample having strong antiferromagnetic interactions, resulting in ferrimagnetic behavior. The frequency-dependent in-phase, χ′(T), and out-of-phase, χ″(T), components of the alternating current (ac) susceptibility can also be used to determine the magnetic ordering temperatures. In a zero applied direct current (dc) field, 1 exhibits a response in both χ′(T) and χ″(T) with a magnetic ordering temperature of 12 K from the initial rise in the χ″(T) data upon cooling (Figure 3). This ordering temperature is consistent with the blocking temperature obtained in the MZFC(T) and MFC(T) data. 1 has an observable frequency dependence, φ ) 0.029, which is attributed to a spin- or cluster-glass behavior.40 Spin- or cluster-glass behavior has been reported for other PBAs and is indicative of a more disordered and/or amorphous material.7,35,41 This is consistent with the PXRD data, where 1 was shown to be amorphous and structurally disordered (vide supra). The temperature at which the maximum in the χ′(T) data becomes frequencydependent corresponds to the temperature at which the

magnetic moments of this disordered material become frozen, and this is termed the freezing temperature, Tf. Tf for 1 was determined [by a peak in the χ′(T) at 33 Hz] to be 11.5 K. The spin- or cluster-glass behavior supports the hypothesis that there are small ferrimagnetic domains separated by a sea of paramagnetic spins (vide supra). At Tf, the spins in these ferrimagnetic domains are randomly frozen and result in an overall net moment, causing an increase in χ (T) and a response in χac(T). M(H) measurements at 2 K show that 1 does not saturate even at 90 kOe, where a magnetization value of 21 000 emu · Oe/mol is observed (Figure S6 in the Supporting Information). At low temperature, the spins in this system (for VIII, S ) 1 and g ) 1.82,7 and for MnII, S ) 5/2 and g ) 2.0036) should be antiferromagnetically aligned, resulting in an expected saturation magnetization value of 31 700 emu · Oe/mol. Hysteresis was also observed for 1 at 2 K (coercive field, Hcr, of 65 Oe) with a remanent magnetization, Mrem, of 730 emu · Oe/mol (Figure 4). 1 is one of the first reported PBA with [VIII(CN)6]3incorporated into the structure, leading to M-NtC-VIII linkages. This material is best described as ferrimagnet, exhibiting a cluster-glass behavior. At 90 K, ferromagnetic spin domains grow in size, resulting in a superparamagneticlike effect with an overall increase in the χT(T) data. When compound 1 is submerged in liquid N2 and removed, the sample becomes attracted to a magnet (displayed in the Table of Contents picture), which is in accordance with the presence of a superparamagnetic-like state above the observed critical temperature of 12 K. As the temperature is decreased further, these spin domains grow until bulk magnetic ordering is obtained at 12 K. At this temperature, (40) Mydosh, J. A. In Spin Glasses: An Experimental Introduction; Taylor and Francis: London, 1993; p 67. φ is a parameter indicative of the amount of spin disorder in a material known as spin-glass behavior: φ ) ∆Tmax/[Tmax(∆ log ω)], where ∆Tmax ) difference between the peak maxima of the temperatures at the high and low frequencies, Tmax ) peak maximum of the temperature at low frequency, and ∆ log ω ) difference in the logarithms of the high and low frequencies (ω). (41) (a) Buschmann, W. E.; Ensling, J.; Gütlich, P.; Miller, J. S. Chem.sEur. J. 1999, 5, 3019. (b) Sendek, M.; Csach, K.; Kavecansky, V.; Lukácová, M.; Marysko, M.; Mitróová, Z.; Zentko, A. Phys. Status Solidi A 2003, 196, 225.

Inorganic Chemistry, Vol. 47, No. 7, 2008

2531

Nelson and Miller

Figure 5. χT(T) (closed symbols) and χ-1(T) (open symbols) for 2 (circles), 3 (squares), and 4 (diamonds). Fits to the Curie–Weiss law, χ ∝ (T - θ)-1, are displayed with solid lines. Samples were cooled in a zero applied field to 5 K, and data were measured upon warming in a 500 Oe applied field.

the spin domains become frozen (Tf ) 11.5 K), resulting in the observed χac response. Compound 2. 2 has a room temperature χT value of 2.20 emu · K/mol that is only slightly lower than the spin-only value of 2.66 emu · K/mol for VIII, S ) 1 and g ) 1.82,7 and for FeII, S ) 2 and g ) 2.21 (Figure 5).42a If linkage isomerization does not occur, then 2 would be formulated as FeII1.5[VIII(CN)6] · 0.85MeCN and all FeII ions would be in a high-spin environment. This would lead to a spin-only χT value at room temperature of 6.32 emu · K/mol for VIII, S ) 1 and g ) 1.82,7 and for FeII, S ) 2 and g ) 2.21.42a This spin-only χT(T) value drastically exceeds the observed value of 2.20 emu · K/mol, indicative that linkage isomerization has occurred and 2’s correct formulation is FeII0.5VIII[FeII(CN)6] · 0.85MeCN. Above 50 K, χ(T ′) could be fit to the Curie–Weiss expression χ ∝ (T - θ)-1 with θ ) -6 K, suggesting that antiferromagnetic coupling dominates the short-range exchange. This is in accordance with reported literature ranges (-8 to +23 K) for other hexacyanoferrate(II) PBAs.43 The shape of the χT(T) data is consistent with 2 being paramagnetic, with no long-range order being observed. Using the θ value of -6 K, χT(T - θ)calc was calculated to be 2.61 emu · K/mol, which is consistent with the observed room temperature χT value, confirming 2’s formulation of FeII0.5VIII[FeII(CN)6] · 0.85MeCN. The 2 K M(H) data display that 2 does not saturate at 50 kOe, where a magnetization of 8000 emu · Oe/mol is observed (Figure S6 in the Supporting Information). The shape of the M(H) curve and the inability to saturate are indicative of a paramagnetic material. In summary, 2 is a paramagnetic material with weak antiferromagnetic coupling (θ ) -6 K). (42) (a) Boudreaux, E. A.; Mulay, L. N. Theory and Applications of Molecular Paramagnetism; Wiley-Interscience: New York, 1976; p 203. (b) Boudreaux, E. A.; Mulay, L. N. Theory and Applications of Molecular Paramagnetism; Wiley-Interscience: New York, 1976; p 214. (43) (a) Egan, L.; Kamenev, K.; Papanikolaou, D.; Takabayashi, Y.; Margadonna, S. J. Am. Chem. Soc. 2006, 128, 6034. (b) MartínezGarcia, R.; Knobel, M.; Reguera, E. J. Phys. Chem. B 2006, 110, 7296.

2532 Inorganic Chemistry, Vol. 47, No. 7, 2008

The absence of bulk magnetic ordering is likely caused by the presence of the diamagnetic [FeII(CN)6]4- (S ) 0) bridging the paramagnetic ions (VIII and FeII), resulting in the distance between paramagnetic sites being greater than 10 Å. This large distance causes no correlation of the spins throughout the bulk materials. Compound 3. 3 has a room temperature χT value of 2.35 emu · K/mol. From the discussion of the IR data above, 3 undergoes a partial linkage isomerization, resulting in the presence of two types of bridging cyanides [CoII-NtC-VIII (2121 cm-1) and CoIII-CtN-VII (2161 cm-1)]. However, the ratio of these linkages cannot be obtained directly from the intensities of the IR absorptions. In an attempt to elucidate the ratio of these linkages, the calculated χT values of the two limiting compositions are investigated (a ) 0 and a ) 1). If a ) 0, then the formulation of 3 is (NEt4)0.10CoII1.5[VIII(CN)6](BF4)0.10 · 0.35MeCN. With this formulation, a spin-only value of 5.58 emu · K/mol (for VIII, S ) 1 and g ) 1.82,7 and for CoII, S ) 3/2 and g ) 2.6042b) is calculated. This calculated χT value is much higher than the observed room temperature value of 2.35 emu · K/mol; thus, this formulation is unlikely. If a ) 1, then the formulation of 3 is (NEt4)0.10CoII0.5VII[CoIII(CN)6](BF4)0.10 · 0.35MeCN. This formulation results in a spin-only value of 3.14 emu · K/mol (for VII, S ) 3/2 and g ) 1.82;7 for CoII, S ) 3/2 and g ) 2.60;42b and for diamagnetic [CoIII(CN)6]3-, S ) 0), which is more consistent with the observed room temperature value of 2.35 emu · K/mol. Above 50 K, χ (T) could be fit to the Curie–Weiss expression χ ∝ (T - θ)-1 with θ ) -15 K, suggesting that antiferromagnetic coupling dominates the short-range exchange (Figure 5), and this θ value is consistent with the literature values of -31 K reported for CoII3[CoII(CN)5]2.25 Using this θ value (-15 K), χT(T - θ)calc was determined to be 5.31 and 2.99 emu · K/ mol for a ) 0 and 1, respectively. In the latter case (a ) 1), no CoII-NtC-VIII linkages remain that would result in the no observable νCN absorption assigned to these linkages (2121 cm-1). This is not the case for 3, where the 2121 cm-1 absorption is slightly more intense than the 2161 cm-1 absorption (the observed ratio of the CoIII-CtN-VII: CoII-NtC-VIII linkages is 0.93). If the νCN absorptions for these two stretching modes had equivalent intensities, the complete composition could be obtained directly from this ratio. The actual composition of 3 is still somewhat of a mystery; however, from the room temperature χT value, it is clear that the formulation exists somewhere between these two stoichiometric limits and is closer to the latter formulation of (NEt4)0.10CoII0.5VII[CoIII(CN)6](BF4)0.10 · 0.35MeCN than the former one. Because the exact formulation could not be obtained from these studies, a generic composition of (NEt4)0.10CoII(1.5-a)VIIa[CoIII(CN)6]a[VIII(CN)6]1-a(BF4)0.10 · 0.35MeCN will be used to describe this material.44 The 2 K M(H) data show that 3 does not saturate at 90 kOe, where a magnetization value of 10 800 emu · Oe/mol is observed (Figure S6 in the Supporting Information). The (44) Another method of determining the oxidation states of the Co and V metal ions in 3 is via X-ray photoelectron spectroscopy; however, at this time, we do not have the capability to perform these measurements.

Incorporation of Substitutionally Labile [VIII(CN)6]3-

shape of the M(H) curve and the inability to saturate are indicative of a paramagnetic material. Hence, 3 is a paramagnet with moderate antiferromagnetic coupling (θ ) -15 K). The chemical formulation is unclear, but a generic formulation of (NEt4)0.10CoII1.5-aVIIa[CoIII(CN)6]a[VIII(CN)6]1-a(BF4)0.10 · 0.35MeCN (a < 1) can be used to represent the binding motifs in this new material. The absence of bulk magnetic ordering is likely caused by the inherent structural disorder, presence of diamagnetic [CoIII(CN)6]3-, and number of different paramagnetic species in 3. Compound 4. 4 has a room temperature χT value of 0.95 emu · K/mol that is only slightly larger than the spin-only value of 0.83 emu · K/mol for S ) 1 and g ) 1.827 for VIII and S ) 0 for NiII. The room temperature χT value confirms the presence of diamagnetic square-planar [NiII(CN)4]2-, which was formed upon linkage isomerization of the cyanide ligand. Above 5 K, χ(T) could be fit to the Curie–Weiss expression (Figure 5) with θ ) -3 K and g ) 1.96, suggesting that antiferromagnetic coupling dominates the short-range exchange. Using this θ value, χT(T - θ)calc was determined to be 0.82 emu · K/mol. In NiII(CN)2 · 1.5H2O31 and [CuII(HL)]2[NiII(CN)4] (where H2L ) 3,9-dimethyl-4,8diazaundeca-4,8-diene-2,10-dione dioxime),45 similar magnetic data are reported, where both compounds contain diamagnetic [NiII(CN)4]2- (S ) 0) and a second paramagnetic metal ion, octahedral NiII (S ) 1) in the former and squareplanar CuII (S ) 1/2) in the latter. The magnetic data for both compounds are representative of the paramagnetic metal sites in these bimetallic systems. The 2 K M(H) data show that 4 does not saturate at 90 kOe, where a magnetization value of 7180 emu · Oe/mol is observed (Figure S6 in the Supporting Information). The shape of the M(H) curve and the inability to saturate are indicative of a paramagnetic material. Thus, 4 is a paramag(45) Ray, A.; Dutta, D.; Mondal, P. C.; Sheldrick, W. S.; Mayer-Figge, H.; Ali, M. Polyhedron 2007, 26, 1012.

netic material with weak antiferromagnetic coupling (θ ) -3 K), and its formulation as (NEt4)0.20VIII[NiII(CN)4]1.6 · 0.10MeCN fits well for a paramagnetic VIII (S ) 1) with g ) 1.96. The absence of bulk magnetic ordering is a result of diamagnetic [NiII(CN)4]2- sites creating a larger separation between paramagnetic sites. Conclusion Incorporation of [VIII(CN)6]3- into PB-structured materials has proven to be synthetically challenging. The reaction of [VIII(CN)6]3- with [MII(NCMe)6]2+ [M ) Cr,7 Mn (1), Fe (2), Co (3), and Ni (4)] in a MeCN solution results in new PB-type materials; however, only one compound in this series maintains the [VIII(CN)6]3- complex after the synthesis is performed (M ) Mn). All of the other compounds in this series have either electron transfer (M ) Cr, Co) and/or partial (M ) Co) or complete (M ) Fe, Ni) linkage isomerization of the cyanide ligand, resulting in the formation of M-CtN-V linkages. The three compounds exhibiting linkage isomerization (M ) Fe, Co, Ni) are paramagnetic down to low temperatures with only moderate (θ ) -15 K for Co) to weak (θ ) -3 K for Ni) short-range antiferromagnetic interactions. When M ) Mn, a ferrimagnetic cluster-glass material is formed that magnetically orders at 12 K (via ac measurements). At 2 K, this compound has an observable coercivity of 65 Oe and remanent magnetization of 730 emu · Oe/mol. Acknowledgment. The authors gratefully acknowledge preliminary data taken by Ian Giles and Shayla Troff and the continued partial support from the U.S. DOE Basic Energy Sciences (Grant DE FG 03-93ER45504) and the AFOSR (Grant F49620-03-1-0175). Supporting Information Available: Figures S1-S6 are available free of charge via the Internet at http://pubs.acs.org. IC701845P

Inorganic Chemistry, Vol. 47, No. 7, 2008

2533